John,
this is an interesting question. I don't have the answer but I am
asking you to share any advice with the mail-list.
I would suggest that your two models ask the questions:
- Do any of the face conditions (i.e. face 1 or face 2...or face 4)
differ from all the object conditions?
- Do all of the face conditions (i.e. face 1 and face 2...and face 4)
applying equal weight differ from all the object conditions?
But don't take my comment for granted. I am not an expert.
I would be particularly curious how you would implement this
experiment in a second level analysis.
Michael Froelich, MD MS
E-mail: [log in to unmask]
office: (205) 975-0145
On Jun 4, 2009, at 5:02 PM, John Gelburg wrote:
> Hi all,
>
> The examples of F-contrast usually show it for one session. I would
> just want to make sure that I am OK with defining it for multiple
> sessions.
> Attached experiment design with 8 conditions and 6 sessions. Four
> first conditions are different face conditions, whereas 4 remaining
> conditions are object conditions. I am interested to see all the
> regions, where activation for faces differed from "0" (F contrast).
> Does it look OK?
>
> I defined it as:
> [1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0
> 0 0 0 0 0 0 1 0 0 0 0 0 0 0]
> [0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1
> 0 0 0 0 0 0 0 1 0 0 0 0 0 0]
> [0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
> 1 0 0 0 0 0 0 0 1 0 0 0 0 0]
> [0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0
> 0 1 0 0 0 0 0 0 0 1 0 0 0 0]
>
> One more thing: if I define my contrast as simple one line I got
> much higher F-values. But is it less correct to make such a grand
> average?
> [1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1
> 1 1 0 0 0 0 1 1 1 1 0 0 0 0]
>
>
> Thanks,
> John
> <F-contrast_multi_sessions.JPG>
|